Add jdbc sink support for Structured Streaming

Currently, spark sql support read and write to jdbc in batch mode, but do not support for Structured Streaming. During work, even thought we can write to jdbc using foreach sink, but providing a more easier way for writing to jdbc would be helpful.